Population Genetics of an Opportunistic Parasitoid in an Agricultural Landscape

Heredity 80 (1998) 152–162 Received 15 November 1996

Population genetics of an opportunistic parasitoid in an agricultural landscape

TY T. VAUGHN* & MICHAEL F. ANTOLIN Department of Biology, Colorado State University, Fort Collins, CO 80523, U.S.A.

Parasitoid insects that use different hosts can have a subdivided population structure that corresponds to host use. A subdivided population structure may favour local adaptation of subpopulations to small-scale environmental differences and may promote their genetic diver- gence. In this paper, heritable Random Ampliﬁed Polymorphic DNA (RAPD) markers visual- ized by single strand conformational polymorphisms (SSCP) analysis were used to examine the population structure of the parasitoid wasp Diaeretiella rapae (Hymenoptera: Braconidae) in an environment where two aphid hosts are available for oviposition. We found 11 codominant and 34 dominant RAPD polymorphisms that conformed to Mendelian segregation patterns. A nested analysis of variance indicated extensive genetic differentiation among six populations of D. rapae that were sampled for two years. Effective migration rates (Nm) between populations ranged from 1.2–1.6 per year, indicating a relatively low dispersal rate. Genetic distances were also calculated between populations and the resulting trees indicated that populations less than 1.0 km from each other were genetically differentiated. Our results indicate that D. rapae populations are genetically subdivided on a small spatial scale that corresponds to host-use patterns.

Keywords: biological control, Diaeretiella rapae, gene ﬂow, parasitoid, population structure, RAPD-PCR.

Introduction Identifying the forces responsible for genetic differentiation of populations is of interest to evolu- Nearly all species of plants and animals display tionary biologists; it is also important to practition- genetic differentiation at some geographical scale, ers of biological control. Biological control theory usually as a result of habitat discontinuity, habitat predicts that useful natural enemies will have high fidelity or limited dispersal. In some cases, local dispersal rates compared to their prey or hosts, environmental variation may cause natural selection especially in heterogeneous environments (van to operate differently among populations, leading to Emden, 1974; Murdoch, 1990; Roderick, 1992). This genetic differentiation (Via, 1994). The evolutionary implies that the most successful natural enemies forces responsible for creating genetic differentia- should not form local subpopulations but should tion among populations are limited migration, small remain panmictic. Despite the recognized import- local population sizes, restricted matings and selec- ance of natural enemy mobility, little is known about tion within a population (Endler, 1986; Hartl & their population structure or about the rates of gene Clark, 1989; Rank, 1992; Roderick, 1996). Dispersal flow between local populations of predators and of individuals into other populations can ameliorate parasitoids (Hopper et al., 1993; Roderick, 1996). these effects by increasing gene flow among popula- One of the problems in the analysis of population tions (Slatkin, 1987, 1994). Therefore, the presence structure is determining the amount of gene flow of population structuring indicates that one or more between populations. It is usually impossible to of these forces may be operating. detect individual insects moving long distances (but see Antolin & Strong, 1987), therefore we must rely *Correspondence: Department of Anatomy and Neurobiology, Washington University School of Medicine, 660 South Euclid on indirect measures such as DNA markers to infer Avenue, St. Louis, MO 63110, U.S.A. E-mail: gene flow (Roderick, 1996). Historically, the major- [email protected] ity of gene flow analyses have examined protein GENETIC POPULATION STRUCTURE OF A PARASITOID WASP 153 polymorphisms (allozyme markers) to describe allele pupates within the aphid host and emerges from the frequency variation in time and space. However, mummy as an adult. Under laboratory conditions allozymes may not provide adequate resolution for (20°C; 60 per cent RH), egg to adult development determining migration patterns because of low requires :9 days. mutation rates and selection acting to limit varia- In northern Colorado, two aphid species are bility of protein allozymes (Avise, 1994). In addition, commonly attacked by D. rapae, the cabbage aphid Hymenoptera have exceptionally low allozyme varia- (Brevicoryne brassicae) and Russian wheat aphid. bility (Graur, 1985). Recent developments of molec- Both aphid species are serious pests to agricultural ular genetic markers that detect DNA-level crops: B. brassicae feeds exclusively on crucifer crops variability have increased the power of insect popu- and D. noxia feeds exclusively on cultivated cereals. lation genetic studies (Avise, 1994; Haymer, 1994; Crucifer crops and cereal crops are grown in close Roderick, 1996). We have developed and enhanced proximity to each other in some regions whereas in the use of Randomly Amplified Polymorphic DNA other regions, cereal crops, mainly dryland wheat, amplified by the Polymerase Chain Reaction are grown in extensive monoculture. This scenario (RAPD-PCR) to generate repeatable and heritable affords D. rapae the opportunity to parasitize two markers useful in molecular taxonomy, genome aphid hosts which may be separated by many kilo- mapping and population genetics (Hiss et al., 1994; metres or as little as several metres. Antolin et al., 1996a,b). In this study, RAPD-PCR markers were used in Field collections conjunction with single strand conformation polymorphism (SSCP) analysis to examine the popula- Wasps were collected from crucifer vegetable crops tion structure of the parasitoid wasp Diaeretiella and from cultivated cereal crops in northern Colo- rapae M’Intosh (Hymenoptera: Braconidae). Diaer- rado. Individual wasps were removed from aphid etiella rapae is a parasitoid of many aphid species colonies on host plants by gently brushing the that feed on various plant types but it is commonly mummified aphids individually into gelatin capsules. found parasitizing aphids that feed on cruciferous Aphid mummies were collected rather than adult plants. In previous studies, we found that D. rapae is wasps to ensure that the wasps analysed were the primarily attracted to cruciferous plant odours progeny of adult females that located and chose to (Vaughn et al., 1996). However, D. rapae is used as parasitize a particular aphid in a particular location. a biological control agent against the Russian wheat Aphid mummies were returned to the laboratory, aphid (Diuraphis noxia) which feeds on cultivated were maintained at 23°C until the adults emerged, grasses in the western United States. Thus, D. rapae then were sexed and frozen at 80°C. has a choice between two aphid hosts where these A hierarchical sampling design was used to collect plant types are grown together. To assess whether D. rapae populations from six fields within three this situation may lead to population differentiation regions of northern Colorado (Table 1). Within each we examined the genetic structure of this parasitoid region, we located two aphid-infested fields less than on a microgeographical scale within managed agri- 1.0 km from each other. Region I was located at cultural systems in northern Colorado. Piedmont Farms, an organic vegetable production area. The two wasp populations collected from this area were sampled from Brussels sprouts and winter Materials and methods wheat, separated by 0.25 km. Region II was located 23 km south-east of Region I. The populations Study organisms sampled in this region were located on 4 V Farms, a Diaeretiella rapae is a solitary endoparasitoid of cereal production farm in Weld County. Both D. many aphid species (Vater, 1971; Stary, 1976) with rapae populations were collected from winter wheat adult females laying a single egg directly into an fields separated by 0.50 km. Region III was located aphid’s body. The parasitized aphid continues to 12 km south-east of Region I and 18 km south-west feed for 3–4 days and typically remains in the same of Region II. The two populations collected in this location after being attacked. Diaeretiella rapae region were 1.0 km apart and were collected from a progresses through four larval instars as it feeds on winter wheat field and a cabbage field. the internal tissues of the aphid, eventually killing it. Each of the six populations was sampled three When the host dies it becomes a mummy consisting times per year (June, July and August) in 1994 and of the hardened exoskeleton of the aphid. The wasp 1995 to test for seasonal differences in population

Table 1 Sampling locations for Diaeretiella rapae populations in Colorado in 1995 and 1996

Population Distance from Distance between ﬁelds Region location Field type region I (km) within each region (km)

I Piedmont Farms Wheat 0 I Piedmont Farms Crucifer 0 0.25 II Weld County Wheat 23 II Weld County Wheat 23 0.05 III Hort Farms Crucifer 13 III BARI Wheat 12 1.0

structure. For each sampling period, 46 individuals 15 h. The gels were subsequently stained using silver were analysed (23 males and 23 females). The field to detect the mobility of the different DNA markers. sampling procedure followed a modified version of Quick Sample, a tool for determining Russian wheat aphid infestations in cereal fields (Legg et al., 1991). Segregation analysis of RAPD-SSCP markers Fields were entered at one corner and a plant was The majority of RAPD markers are expressed as sampled for D. rapae mummies every 10–20 paces. phenotypically dominant markers in diploid organ- Only one wasp mummy was collected from any isms. However, because of the high resolution plant. This procedure ensured that the entire field offered by the PAGE technique, SSCP procedures, was sampled and reduced the chance of collecting and the haplo-diploid sex determination in Hy- several progeny from a single female wasp. menoptera, we have also identified numerous codominant markers using a segregation analysis. Inherit- ance of the bands was tested by generating full-sib DNA extraction, RAPD-SSCP and electrophoresis families from each of the field populations. Two DNA was isolated from individual wasps following families were created for each of the two Piedmont the protocol of Coen et al. (1982). DNA was resus- populations and one family was created for the pended in 100 mL TE (10 mM TRIS-HCl, 1 mM remaining four populations. Diaeretiella rapae EDTA, pH 8.0) and 1 mL of DNA template was families were produced by collecting individuals as used in each 50 mL PCR reaction. Details of the mummies and rearing them in an incubator (25°C, RAPD-PCR protocols can be found in Black et al. 60 per cent RH). Upon emergence, mated pairs (1992) and details of the SSCP (Single Strand were formed and the females allowed to oviposit on Conformational Polymorphisms) analysis can be an aphid-infested plant for 24 h. The mated pair was found in Hiss et al. (1994) and Antolin et al. (1996b). then stored at 80°C. The progeny were individ- Each set of PCR reactions was checked for contami- ually collected as mummies, allowed to emerge as nation by using a negative control (all reagents adults and also stored at 80°C. These families except template DNA). Samples were stored at 4°C were screened with 18 RAPD primers. We chose the until being electrophoresed. following four primers from Operon Technologies: ,Complete details of the polyacrylamide gel A05, C01, C04 and BAMH1 (5؅-ATGGATCCGC-3؅ electrophoresis (PAGE) procedure are described in designed by N. M. DuTeau, Colorado State Univer- Hiss et al. (1994) and Black & DuTeau (1996). sity) with which to test Mendelian inheritance Modifications to this procedure are in Antolin et al. patterns. Repeatability was then tested by amplifying (1996b). Briefly, PCR products (5 mL of a 50 mL the DNA of each individual in each family twice. sample) were mixed with 2 mL loading buffer at Amplified markers were scored directly from room temperature. The products were electrophor- dried gels by measuring band mobility relative to esed on large (40Å50 cm), thin (0.4 mm) glycerol (5 known size markers (1 kb ladder, BRL Labora- per cent) polyacrylamide (5 per cent, 2 per cent tories). To estimate sizes of the amplified DNA frag- cross-linking) gels. Samples (5 mL) were loaded into ments, size standards were fitted to an inverse 6 mm sharkstooth combs. Electrophoresis proceeded function that relates fragment size and mobility at constant voltage (350 V) at room temperature for (Schaffer & Sederoff, 1981). Each marker was

© The Genetical Society of Great Britain, Heredity, 80, 152–162. GENETIC POPULATION STRUCTURE OF A PARASITOID WASP 155 labelled by the primer used in the PCR reaction and Genetic distance by its mobility on the gel (e.g. A05.275). The second method of analysis was based upon pair- wise genetic distances between populations, which Analysis of population subdivision does not depend upon the assumption of population equilibrium (Slatkin, 1993). The number of markers The RAPD-PCR markers generated from our ﬁeld- generated with RAPD-PCR allowed us not only to collected populations were analysed in two ways. estimate genetic distances but also to perform robust The most commonly used methods to estimate statistical analyses of population subdivision at population subdivision are Wright’s F-statistics and smaller spatial scales. In conjunction with W. C. related estimates (Wright, 1978; Cockerham & Black IV, a computer program called RAPDDIST Weir, 1993; Slatkin, 1993). These indices measure (Black & Antolin, see Appendix) was developed that the correlation between alleles found in subpopula- uses RAPD data to estimate genetic distances for tions, relative to the population as a whole. haploid and diploid organisms. Nei’s unbiased FST represents the standardized variance in allele genetic distances (Nei, 1978) between pairs of popu- frequencies among local populations and ranges lations were then submitted to PHYLIP 3.5c (Felsen- between 0, indicating no genetic differentiation, and stein, 1995) to create cluster diagrams (trees) based 1, indicating complete differentiation. In order to upon UPGMA, an unweighted pair-group method with calculate FST from dominant RAPD markers, it must arithmetic averaging clustering analysis (Avise, be assumed that populations are in Hardy–Wein- 1994). RAPDDIST further allows for statistical testing berg equilibrium. However, by examining the of individual branches on the trees by resampling frequencies of the codominant alleles from the the data, thus testing for consistency within the female samples, the assumption of Hardy–Weinberg RAPD data set in forming the observed groupings. could be explicitly tested. We used the computer Each sex was analysed separately for each sampling program BIOSYS-1 (Swofford & Selander, 1981) and a date and also by combining all sampling dates. chi-squared analysis to test for deviations from Hardy–Weinberg equilibrium for each codominant RAPD marker. Results Estimates of F and effective migration rates ST Segregation analysis (Nm) among the females in each population (n = 23) were calculated using the computer The four primers generated a total of 34 dominant program RAPDFST provided by W. C. Black IV (presence/absence) markers and 11 codominant (Department of Microbiology, Colorado State markers in the families, all of which were poly- University, see Appendix). The program calculates morphic and had an observed frequency of less than standard FST and effective migration rates (Wright, 0.92, as recommended by Lynch & Milligan (1994). 1978) as well as estimates corrected for small sample Primer A05 produced 11 dominant and four codo- sizes (Cockerham & Weir, 1993) and for both small minant markers, C01 produced nine dominant and sample sizes and the effects of dominant loci three codominant markers, C04 produced eight commonly detected by RAPD-PCR (Lynch & Milli- dominant and three codominant markers and gan, 1994). BAMH1 produced six dominant and one codominant marker (Table 2). Repeatability was veriﬁed by reamplifying the DNA of each individual in each Linkage disequilibrium family with all four primers a second time and iden- Linkage disequilibrium describes the nonrandom tifying the same markers. The inheritance of each association of alleles (Lewontin & Kojima, 1960) marker conformed to Mendelian segregation and can arise in subdivided populations from patterns among the full-sib families. Figure 1 epistatic natural selection within subpopulations demonstrates common patterns found in the segre- (Lewontin, 1974, pp. 273–318), random genetic drift gation of RAPD-PCR markers. Mendelian inherit- among subpopulations (Ohta & Kimura, 1969) or ance of codominant RAPD markers for haplo- actual linkage among the markers. Di-locus linkage diploid families can be determined as follows: if a disequilibrium was estimated using RAPDLD (Black, female having a fast and slow allele is mated to a 1995; Apostol et al., 1996; see Appendix) which esti- male with a slow allele, the male offspring will have mates linkage disequilibrium among RAPD alleles either a fast or a slow allele and the female offspring in the subpopulations. will be either heterozygous or homozygous for the

© The Genetical Society of Great Britain, Heredity, 80, 152–162. 156 T. T. VAUGHN & M. F. ANTOLIN slow allele (Fig. 1a). In the case of a dominant All three estimates of population subdivision were marker, if a female homozygous for the null allele similar in this study for males and females (no band present) is mated to a male that carries a combined: FST estimated from Wright (1951) was copy of the allele, none of the male offspring will 0.066 in 1994 and 0.073 in 1995 (P0.01); Weir & show the marker but all of the female offspring will Cockerham’s y, which accounts for small sample have a copy because they developed from fertilized sizes, was estimated as 0.063 in 1994 and 0.070 in eggs (Fig. 1b). 1995; and FST estimated from Lynch & Milligan (1994), whose method accounts for small and unequal sample sizes and dominance effects inherent with RAPD loci, was estimated at 0.061 in Estimation of population subdivision 1994 and 0.069 in 1995. FST- and y-values for each Deviations from Hardy–Weinberg equilibrium were sampling date are given in Table 4. tested for each of the 11 codominant loci in each of The data set was subjected to a nested analysis of the six populations using the female wasps. No variance following Wright (1978) using BIOSYS-1 deviations from Hardy–Weinberg equilibrium were (Swofford & Selander, 1981) to determine the level 2 found (all x13.33; P0.10) except for one locus in at which the populations were genetically sub- 2 the population obtained from BARI (x1 = 3.94; divided. Variance components indicate that most of P0.05). However, this was not significant when the the variation within the population occurred variances were adjusted using Bonferroni corrections between fields rather than between the areas (P0.13). The mean heterozygosity, averaged across containing the fields for all sampling dates in each sample dates and RAPD loci for each population year (Table 5). The effective migration rates was calculated and ranged from 0.335–0.443 obtained for all populations for 1994 and 1995 were (Table 3). quite low at 1.2 and 1.6 wasps per year, respectively.

Table 2 Frequency of RAPD markers for all six Diaeretiella rapae populations for females and males for each year. Shown are the frequencies for the band-present alleles for the dominant markers and the ‘fast’ alleles for the codominant markers (denoted by *)

1994 1995 1994 1995

Marker Female Male Female Male Marker Female Male Female Male

A05.275 0.861 0.842 0.755 0.701 C01.372 0.667 0.690 0.717 0.966 A05.320* 0.694 0.667 0.761 0.696 C01.415 0.283 0.370 0.304 0.396 A05.380 0.619 0.671 0.543 0.630 C01.425 0.333 0.310 0.239 0.283 A05.465 0.521 0.474 0.590 0.523 C01.452 0.543 0.652 0.478 0.578 A05.486 0.385 0.425 0.467 0.511 C01.569 0.696 0.743 0.657 0.701 A05.496 0.306 0.286 0.304 0.413 C01.597 0.667 0.723 0.698 0.723 A05.585 0.717 0.697 0.755 0.722 C01.635* 0.583 0.501 0.555 0.497 A05.620 0.239 0.282 0.360 0.392 C01.791 0.534 0.611 0.567 0.623 A05.682* 0.861 0.810 0.752 0.796 C01.815* 0.457 0.399 0.489 0.409 A05.761* 0.500 0.432 0.564 0.499 C01.1217* 0.472 0.569 0.512 0.577 A05.820 0.417 0.476 0.289 0.334 C01.1615 0.389 0.443 0.355 0.452 A05.855* 0.652 0.718 0.578 0.591 C01.2376 0.826 0.761 0.773 0.697 A05.894 0.804 0.761 0.891 0.800 C04.289 0.306 0.286 0.327 0.291 A05.975 0.572 0.611 0.609 0.701 C04.351 0.778 0.786 0.696 0.775 A05.1254 0.279 0.299 0.317 0.342 C04.415* 0.472 0.515 0.505 0.543 BAMH1.331 0.196 0.174 0.231 0.202 C04.496 0.217 0.330 0.256 0.342 BAMH1.415 0.239 0.196 0.341 0.225 C04.544 0.348 0.477 0.389 0.514 BAMH1.497 0.712 0.784 0.754 0.798 C04.565 0.696 0.772 0.819 0.832 BAMH1.532 0.543 0.652 0.603 0.687 C04.705* 0.286 0.378 0.396 0.403 BAMH1.615 0.457 0.340 0.514 0.448 C04.967 0.222 0.298 0.201 0.279 BAMH1.732* 0.513 0.599 0.609 0.623 C04.1170* 0.592 0.501 0.611 0.577 BAMH1.825 0.876 0.778 0.765 0.698 C04.1852 0.761 0.744 0.679 0.778 C04.2005 0.577 0.523 0.613 0.549

Fig. 1 Examples of silver-stained PAGE gels demonstrating the inheritance patterns of (a) codominant RAPD-PCR markers and (b) dominant RAPD-PCR markers in Diaeretiella rapae. See text for details of haplo-diploid Mendelian inheritance patterns.

Linkage disequilibrium Cluster analysis With 34 RAPD loci there were 561 [(34Å33)/2] Cluster diagrams were generated for each sex for pair-wise comparisons made to estimate linkage each sample date and for each sex for all sample disequilibrium. The amount of linkage disequi- dates combined for each year. There were no differ- librium detected in any population was less than 3 ences between the trees generated for males or per cent and in no population was this amount signi- females at any sampling date and there were no ﬁcant at a = 0.05. This analysis indicates that the signiﬁcant differences in topology between sample RAPD markers used in this study were not linked dates within either sex using strict consensus (50 with one another. per cent grouping). Trees generated using females from all sampling dates within 1994 and 1995 are shown in Fig. 2. Although there is weak support for population subdivision among areas (e.g. Piedmont Table 3 Mean heterozygosity for the 11 codominant alleles (two alleles per locus) averaged across sample y dates and RAPD loci for each population of Diaeretiella Table 4 Estimates of FST and for Diaeretiella rapae by rapae. Numbers in parentheses indicate standard errors date for 1994 and 1995 y Population Heterozygosity FST (W) FST (L & M) Sampling Piedmont wheat 0.335 (0.045) date 1994 1995 1994 1995 1994 1995 Piedmont crucifer 0.416 (0.024) Weld wheat I 0.443 (0.030) June 0.070 0.065 0.060 0.068 0.064 0.071 Weld wheat II 0.425 (0.039) July 0.059 0.071 0.058 0.062 0.056 0.063 Hort Farms crucifer 0.392 (0.024) August 0.066 0.074 0.070 0.073 0.061 0.069 BARI wheat 0.412 (0.055) (W), Wright; (L & M), Lynch & Milligan.

Table 5 Variance components combined across loci from the nested analysis of variance for Diaeretiella rapae populations collected in Colorado during 1994 and 1995

Variance components Sampling date Comparison 1994 1995

June Fields within regions 0.619 0.613 Fields within the total 0.424 0.501 Regions within the total 0.035 0.059 July Fields within regions 0.771 0.690 Fields within the total 0.301 0.405 Regions within the total 0.070 0.013 August Fields within regions 0.809 0.771 Fields within the total 0.299 0.316 Regions within the total 0.094 0.073

Farms wheat and crucifer populations vs. Weld tered by parasitoids in agricultural environments County wheat populations, Fig. 2a), complete may cause population differentiation. support is given only to each branch that includes This study of the population genetic structure of the same fields in two different years except for D. rapae is one of the first studies to demonstrate Weld County populations (Fig. 2b). The Weld genetic differentiation in a parasitoid or biological County populations group together within the same control agent. Because of its use as a biocontrol year for both 1994 and 1995. Interestingly, crucifer agent, the biology of this wasp has been extensively fields (Piedmont Farms and Hort Farms) remained studied. However, most of the research conducted genetically distinct from nearby wheat fields (Pied- on this wasp has concentrated on how it locates mont Farms and BARI), although they were less aphid hosts (Read et al., 1970; Ayal, 1987; Sheehan than 1.0 km from each other. & Shelton, 1989b), especially with reference to chemical cues (Sheehan & Shelton, 1989a; Reed et al., 1991; Vaughn et al., 1996). Diaeretiella rapae is Discussion attracted to the volatiles emitted from crucifer The habitats of most insect species are hetero- plants and may possess specialized receptors used in geneous and discontinuous in time and space and locating habitats with these plant volatiles. Yet, D. typically contain patches that provide essential rapae parasitizes many other aphid species not found resources such as food or reproduction sites (Elton, on crucifer plants. In this study, the population 1949; Southwood, 1977). In such cases, gene flow structure of D. rapae was examined with particular between populations can be restricted because indi- reference to habitat types which occur in diverse viduals tend to move mainly within physically or agricultural systems. otherwise separated habitats that contain essential One of the difficulties encountered in population resources. In many invertebrates, a combination of genetic studies, especially of Hymenoptera, is the limited dispersal and habitat fidelity results in level of resolution offered by indirect estimates of marked population subdivision, sometimes even over gene flow. With the advent of RAPD-PCR tech- small geographical distances. For example, popula- nology, it is now possible to estimate differentiation tion substructuring on a microgeographical scale has of populations on a relatively small spatial scale been demonstrated in a wide variety of organisms, because of the large number of genetic polymorph- showing that gene flow is often extremely limited, isms. In addition, the use of large polyacrylamide even over relatively short distances (Bilton, 1992; gels combined with the greater resolution of silver Rank, 1992; Seppa & Pamilo, 1995; Apostol et al., staining made the analysis of these polymorphisms 1996; West, 1996). Resource fragmentation may be far easier and more accurate than other methods of common in agricultural systems, especially in para- detecting genetic differences between populations. If sitoid–host interactions where more than one host the inheritance patterns of RAPD-PCR markers can type is available. Thus, the diverse habitat encoun- be established as we have shown in this study, these

© The Genetical Society of Great Britain, Heredity, 80, 152–162. GENETIC POPULATION STRUCTURE OF A PARASITOID WASP 159 markers become as effective as any other genetic Lynch & Milligan (1994) all produced similar esti- marker used in population-level studies (Carlson et mates of population differentiation. Furthermore, al., 1991; Hunt & Page, 1992; Apostol et al., 1993; the FST-values and variance components indicate Fondrk et al., 1993; Mitchell-Olds, 1995; Antolin et that most of the variation in northern Colorado al., 1996a,b). The use of genetic markers such as occurs between fields rather than between areas RAPDs also tests whether the entire genome containing the fields, for all sampling dates within supports the observed groupings rather than just a each year. Thus, the individual fields within an area few loci. This is further supported because none of are as different from each other as are fields in the RAPD markers was in significant linkage different areas, even though pairs of fields within disequilibrium with another. any area are separated by 1 km or less and areas are The estimates of population subdivision via separated by as much as 23 km (Fig. 2b). In addi- Wright’s F-statistics certainly indicate that the D. tion, the comparisons of sampling dates within each rapae population is subdivided. The methods of year indicate that the subdivision among populations Wright (1951), Weir & Cockerham (1984) and is maintained from year to year, even in small

Fig. 2 Cluster diagrams showing the structure of Diaeretiella rapae female populations in Colorado. Shown are Nei’s standard genetic distances among populations of individuals collected during August 1994 and 1995. At each branching point is indicated the bootstrap support (out of 100) for that branch. (a) The pattern indicates that the population of D. rapae in northern Colorado is subdivided into many smaller populations, probably a result of limited dispersal, even between fields that are less than 1.0 km from each other (Piedmont wheat fields vs. Piedmont crucifer fields). (b) After collapsing the branching points with low support (fewer than 75 times out of 100), the pattern still indicates that fields are different from one another.

© The Genetical Society of Great Britain, Heredity, 80, 152–162. 160 T. T. VAUGHN & M. F. ANTOLIN geographical areas, and does not depend on tion between local patches may cause conflicts that recolonization each year. produce ‘trade-offs’ in performance unless the popu- The analysis of the genetic distances between lations can adapt to different environments (Via & populations also demonstrates that the population is Lande, 1985). The possibility of host adaptation is structured in the agricultural environment. These currently being investigated in these D. rapae results indicate that once wasps locate habitats with populations. aphids, they remain in those areas and disperse less than 1 km, whether those habitats are wheat fields or crucifer fields. However, the two populations Acknowledgements from the Weld County wheat fields appear to be The authors wish to thank Piedmont Organic Farms more homogeneous (Fig. 2). The two fields group for their co-operation and financial support for this within the same year for both years and the internal project. We also thank Dr Whitney Cranshaw for node grouping the Weld County Region is high. access to field sites. The manuscript benefitted from This has important implications for biological comments by Dr Joan Herbers and two anonymous control programmes where wasps are released each reviewers. This research was funded by The Organic year. For example, if there are no ameliorating Farming Research Foundation and Colorado Agri- factors involved, D. rapae will most likely remain in cultural Experiment Station grant no. 697. the specific fields where the releases occur and have low migration rates between fields. Indeed, field observations and previous work with D. rapae in References flight tunnels (Sheehan & Shelton, 1989a; Vaughn et ANTOLIN, M. F. AND STRONG, D. R. 1987. Long-distance al., 1996), showed that these wasps have a low dispersal by a parasitoid (Anagrus delicatus, Mymari- tenacity for sustained flight, and spend most of their dae) and its host. Oecologia, 73, 288–292. time at or below the vegetation canopy. ANTOLIN, M. F., GUERTIN, D. S. AND PETERSEN, J. J. 1996a. Although there are numerous studies of herbivo- The origin of gregarious Muscidifurax (Hymenoptera: rous insect populations that demonstrate genetic Pteromalidae) in North America: an analysis using subdivision at small spatial scales (McCauley & molecular markers. Biol. Control, 6, 76–82. Eanes, 1987; McCauley et al., 1988; McPheron et al., ANTOLIN, M. F., BOSIO, C. F., COTTON, J., SWEENEY, W., 1988; Guttman & Weigt, 1989; Guttman et al., 1989; STRAND, M. R. AND BLACK, W. C., IV. 1996b. Intensive Rank, 1992), this is the first study of a parasitoid linkage mapping in a wasp (Bracon hebetor) and a mosquito (Aedes aegypti) with SSCP analysis of RAPD wasp to demonstrate genetic subdivision. Predatory markers. Genetics, 143, 1727–1738. lady beetles (Coleoptera: Coccinellidae) are the only APOSTOL, B. L., BLACK, W. C., IV, REITER, P. AND MILLER, B. natural enemies whose population structure has R. 1996. Population genetics with RAPD–PCR been studied extensively (Krafsur et al., 1992, 1995; markers: the breeding structure of Aedes aegypti in Steiner & Grasela, 1993; Coll et al., 1994). These Puerto Rico. Heredity, 76, 325–334. studies employed allozyme electrophoresis to AVISE, J. C. 1994. Molecular Markers, Natural History and examine gene flow and the main conclusion drawn Evolution. Chapman and Hall, New York. from each study was that the lady beetle population AYAL, Y. 1987. The foraging strategy of Diaeretiella rapae. was not subdivided, even between states in some I. The concept of the elementary unit of foraging. J. cases. This suggests that the high mobility of cocci- Anim. Ecol., 56, 1057–1068. nellids contributes to their effectiveness as biological BILTON, D. T. 1992. Genetic population structure of the postglacial relict diving beetle Hydroporus glabriusculus control agents. A recent study using allozymes, Aube (Coleoptera: Dytiscidae). Heredity, 69, 503–511. however, has demonstrated genetic subdivision BLACK, W. C. 1995. RAPDFST 3. 0 - A FORTRAN program to between Hawaiian populations of the coccinellid estimate Fst and effective migration rates among Curinus coeruleus (Follett & Roderick, 1996). The subpopulations using RAPD-PCR files. Department of levels of gene flow found in this system are similar Microbiology, Colorado State University, Ft. Collins, to those found in our study. CO 80523, U.S.A. For insects that inhabit heterogeneous environ- BLACK, W. C. AND DUTEAU, N. M. 1996. RAPD-PCR and ments, the association with patchily distributed SSCP analysis for insect population genetic studies. In: resources has important consequences for the evolu- Crampton, J., Beard, C. B. and Louis, C. (eds) The tion of insects whose dispersal is limited. Because Molecular Biology of Insect Disease Vectors: A Methods the greatest opportunity for genetic variation in Manual, pp. 361–373. Chapman and Hall, New York. BLACK, W. C., DUTEAU, N. M. PUTERKA, G. J., NECHOLS, J. R. resource use occurs within patchy environments (Via AND PETTORINI, J. M. 1992. Use of random amplified & Lande, 1985; Futuyma & Philippi, 1989), migra-

polymorphic DNA polymerase chain reaction (RAPD- HOPPER, K. R., ROUSH, R. T. AND POWELL, W. 1993. PCR) to detect DNA polymorphisms in aphids (Homo- Management of genetics of biological-control introduc- ptera: Aphididae). Bull. Ent. Res., 82, 151–159. tions. Ann. Rev. Ent., 38, 27–51. CARLSON, J. E., TULSIERAM, L. K., GLAUBITZ, J. C., LUK, V. HUNT, G. J. AND PAGE, R. E. 1992. Patterns of inheritance W. K., KAUFFELDT, C. AND RUTLEDGE, R. 1991. Segrega- with RAPD molecular markers reveal novel type of

tion of random amplified DNA markers in F1 progeny polymorphisms in the honey bee. Theor. Appl. Genet., of conifers. Theor. Appl. Genet., 83, 194–200. 85, 15–20. COCKERHAM, C. C. AND WEIR, B. S. 1993. Estimation of KRAFSUR, E. S., OBRYCKI, J. J. AND FLANDERS, R. V. 1992. gene flow from F-statistics. Evolution, 47, 855–863. Gene flow in populations of the seven-spotted lady COEN, E. S., STRACHAN, T. AND DOVER, G. 1982. Dynamics beetle, Coccinella septempunctata. J. Hered., 83, of concerted evolution of ribosomal DNA and histone 440–444. gene families in the melanogaster species subgroup of KRAFSUR, E. S., OBRYCKI, J. J. AND SCHAEFER, P. W. 1995. Drosophila. J. Mol. Biol., 158, 17–35. Genetic heterozygosity and gene flow in Coleomegilla COLL, M., GARCIA DE MENDOZA, L. AND RODERICK, G. K. maculata De Geer (Coleoptera: Coccinellidae). Biol. 1994. Population structure of a predatory beetle: the Control, 5, 104–111. importance of gene flow for intertrophic level inter- LEGG, D. E., HEIN, G. L. AND PEAIRS, F. B. 1991. Sampling actions. Heredity, 72, 228–236. Russian Wheat Aphid in the Western Great Plains. Great ELTON, C. 1949. Population interspersion: an essay on Plains Agricultural Council, Pub. no. 138. Colorado animal community patterns. J. Ecol., 37, 1–23. State University Cooperative Extension. ENDLER, J. A. 1986. Natural Selection in the Wild. Princeton LEWONTIN, R. C. 1974. The Genetic Basis of Evolutionary University Press, Princeton, NJ. Change. Columbia University Press, New York. FELSENSTEIN, J. 1995. PHYLIP: Phylogeny Inference Package. LEWONTIN, R. C. AND KOJIMA, K. 1960. The evolutionary Version 3.5c. Department of Genetics, Box 357360, dynamics of complex polymorphisms. Evolution, 14, University of Washington, Seattle, WA 98195–7360, 458–472. U.S.A. LYNCH, M. AND MILLIGAN, B. G. 1994. Analysis of popula- FOLLETT, P. A. AND RODERICK, G. K. 1996. Genetic estimate tion genetic structure with RAPD markers. Mol. Ecol., of dispersal ability in the leucaena psyllid predator 3, 91–99. Curinus coeruleus (Coleoptera: Coccinellidae); implica- McCAULEY, D. E. AND EANES, W. F. 1987. Hierarchical tions for biological control. Bull. Ent. Res., 86, 355–361. population structure analysis of the milkweed beetle, FONDRK, M. K., PAGE, R. E., JR. AND HUNT, G. J. 1993. Tetraopes tetraophthalmus (Forster). Heredity, 58, Paternity analysis of worker honeybees using Random 193–201. Amplified polymorphic DNA. Naturwissenschaften, 80, McCAULEY, D. E., WADE, M. J., BREDEN, F. J. AND WOHLT- 226–231. MAN, M. 1988. Spiral and temporal variation in group FUTUYMA, D. J. AND PHILIPPI, T. E. 1989. Genetic variation relatedness: evidence from the imported willow leaf and covariation in response to host plants by Alsophila beetle. Evolution, 42, 184–192. pometaria (Lepidoptera: Geometridae). Evolution, 41, McPHERON, B. A., COURTNEY SMITH, D. AND BERLOCHER, S. 269–279. H. 1988. Genetic differences between host races of GRAUR, D. 1985. Gene diversity in Hymenoptera. Evolu- Rhagoletis pomonella. Nature, 336, 64–66. tion, 39, 190–199. MITCHELL-OLDS, T. 1995. The molecular basis of quantita- GUTTMAN, S. I. AND WEIGT, L. A. 1989. Macrogeographic tive genetic variation in natural populations. Trends genetic variation in the Enchenopa binotata complex Ecol. Evol., 10, 324–328. (Homoptera: Membracidae). Ann. Entomol. Soc. Am., MURDOCH, W. W. 1990. The relevance of pest-enemy 82, 225–231. models to biological control. In: Mackauer, M., Ehler, GUTTMAN, S. I., WILSON, T. AND WEIGT, L. A. 1989. Micro- L. E. and Roland, J. (eds) Critical Issues in Biological geographic variation in the Enchenopa binotata complex Control, pp. 1–24. Intercept Ltd, Andover, U.K. (Homoptera: Membracidae). Ann. Entomol. Soc. Am., NEI, M. 1978. Estimation of average heterozygosity and 82, 156–165. genetic distance from a small number of individuals. HARTL, D. L. AND CLARK, A. G. 1989. Principles of Popula- Genetics, 89, 583–590. tion Genetics, 2nd edn. Sinauer, Sunderland, MA. RANK, N. E. 1992. A hierarchical analysis of genetic differ- HAYMER, D. S. 1994. RAPDs and microsatellites: what are entiation in a montane leaf beetle Chrysomela aenei- they and can they tell us anything we don’t already collis (Coleoptera: Chyromelidae). Evolution, 46, know? Ann. Entomol. Soc. Am., 87, 717–722. 1097–1111. HISS, R. H., NORRIS, D. E., DIETRICH, C. H., WHITCOME, R. F., READ, D. P., FEENY, P. P. AND ROOT, R. B. 1970. Habitat WEST, D. F., BOSIO, C. F. ET AL. 1994. Molecular taxon- selection by the aphid parasite Diaeretiella rapae omy using single-strand conformation polymorphism (Hymenoptera: Braconidae) and hyperparasite Charps (SSCP) analysis of mitochondrial DNA genes. Insect brassicae (Hymenoptera: Cynipidae). Can. Ent., 102, Mol. Biol., 3, 171–182. 1567–1578.

REED, D. K., WEBSTER, J. A. JONES, B. G. AND BURD, J. D. rapae to semiochemicals. Entomologia exp. appl., 78, 1991. Tritrophic relationships of Russian wheat aphid 187–196. (Homoptera: Aphididae), a Hymenopterous parasitoid VIA, S. 1994. The evolution of phenotypic plasticity: what (Diaeretiella rapae M’intosh), and resistant and suscep- do we really know? In: Real, L. A. (ed.) Ecological tible small grains. Biol. Control, 1, 35–41. Genetics, pp. 35–57. Princeton University Press, RODERICK, G. K. 1992. Post-colonization evolution of Princeton, NJ. natural enemies. In: Kauffman, W. C. and Nechols, J. VIA, S. AND LANDE, R. 1985. Genotype-environment inter- R. (eds) Selection Criteria and Ecological Consequences action and the evolution of phenotypic plasticity. Evolu- of Importing Natural Enemies, pp. 71–86. Thomas Say tion, 39, 505–523. Publ., Entomol. Soc. America. WEIR, B. S. AND COCKERHAM, C. C. 1984. Estimating RODERICK, G. K. 1996. Geographic structure of insect F-statistics for the analysis of population structure. populations: gene flow, phylogeography, and their uses. Evolution, 38, 1358–1370. Ann. Rev. Ent., 41, 325–352. WEST, D. F. 1996. Biology of snow-pool Aedes (Diptera: SCHAFFER, H. E. AND SEDEROFF, R. R. 1981. Improved esti- Culicidae) in Larimer County, Colorado. Ph.D. disserta- mation of DNA fragment lengths from agarose gels. tion, Colorado State University. Analyt. Biochem., 115, 113–122. WRIGHT, S. 1951. The genetical structure of populations. SEPPA, P. AND PAMILO, P. 1995. Gene flow and population Ann. Eugen., 15, 323–354. viscosity in Myrmica ants. Heredity, 74, 200–209. WRIGHT, S. 1978. Evolution and Genetics of Populations, SHEEHAN, W. AND SHELTON, A. M. 1989a. The role of vol. 4, Variability Within and Among Natural Populations. experience in plant foraging by the aphid parasitoid University of Chicago Press, Chicago, IL. Diaeretiella rapae (Hymenoptera: Aphidiidae). J. Insect Behav., 2, 743–759. SHEEHAN, W. AND SHELTON, A. M. 1989b. Parasitoid response to concentration of herbivore food plants: Appendix finding and leaving plants. Ecology, 70, 993–998. Information on obtaining FORTRAN programs for SLATKIN, M. 1987. Gene flow and the geographic structure analysis of RAPD-PCR data of natural populations. Science, 236, 787–792. SLATKIN, M. 1993. Isolation by distance in equilibrium and The programs are available by anonymous FTP file non-equilibrium populations. Evolution, 47, 264–279. from Colorado State University and run in batch SLATKIN, M. 1994. Gene flow and population structure. In: mode. Use the following steps to acquire the files via Real, L. A. (ed.) Ecological Genetics, pp. 3–17. electronic mail (computer prompts appear in < > ). Princeton University Press, Princeton, NJ. SOUTHWOOD, T. R. E. 1977. Habitat, the template for 1. Once logged onto your network, type: ecological strategies? J. Anim. Ecol., 46, 337–365. ftp lamar.colostate.edu STARY, P. 1964. Food specificity in the Aphidiidae (Hyme- 2. < user: > anonymous noptera). Entomophaga, 9, 91–99. 3. < password: > your id STEINER, W. W. M. AND GRASELA, J. J. 1993. Population genetics and gene variation in the predator, Coleome- 4. cd pub/wcb4 gilla maculata (de Geer) (Coleoptera: Coccinellidae). 5. binary Ann. Entomol. Soc. Am., 86, 309–321. 6. dir SWOFFORD, D. L. AND SELANDER, R. B. 1981. BIOSYS-1: A You will see rapds.zip. This contains a series of FORTRAN program for the comprehensive analysis of seven zip files that correspond with each of seven electrophoretic data in population genetics and system- RAPD data analysis programs. The source code, atics. J. Hered., 72, 281–283. executable code, a test dataset with output and VAN EMDEN, H. F. 1974. Pest Control and Its Ecology. documentation for each program are included. Edward Arnold, London. 7. To transfer a copy of the files, type: VATER G. VON 1971. Uber ubreitung und orientierung von Diaeretiella rapae (Hymenoptera: Aphididae). Z. Ang. get rapds.zip Entomol., 68, 187–225. 8. Once the programs are on your system, the files VAUGHN, T. T., ANTOLIN, M. F. AND BJOSTAD, L. B. 1996. can be unzipped using PKUNZIP. Behavioral and physiological responses of Diaeretiella